Antibody Immunity to Zika Virus among Young Children in a Flavivirus-Endemic Area in Nicaragua

Abstract

Objective: To understand the dynamics of Zika virus (ZIKV)-specific antibody immunity in children born to mothers in a flavivirus-endemic region during and after the emergence of ZIKV in the Americas. Methods: We performed serologic testing for ZIKV cross-reactive and type-specific IgG in two longitudinal cohorts, which enrolled pregnant women and their children (PW1 and PW2) after the beginning of the ZIKV epidemic in Nicaragua. Quarterly samples from children over their first two years of life and maternal blood samples at birth and at the end of the two-year follow-up period were studied. Results: Most mothers in this dengue-endemic area were flavivirus-immune at enrollment. ZIKV-specific IgG (anti-ZIKV EDIII IgG) was detected in 82 of 102 (80.4%) mothers in cohort PW1 and 89 of 134 (66.4%) mothers in cohort PW2, consistent with extensive transmission observed in Nicaragua during 2016. ZIKV-reactive IgG decayed to undetectable levels by 6-9 months in infants, whereas these antibodies were maintained in mothers at the year two time point. Interestingly, a greater contribution to ZIKV immunity by IgG3 was observed in babies born soon after ZIKV transmission. Finally, 43 of 343 (13%) children exhibited persistent or increasing ZIKV-reactive IgG at ≥9 months, with 10 of 30 (33%) tested demonstrating serologic evidence of incident dengue infection. Conclusions: These data inform our understanding of protective and pathogenic immunity to potential flavivirus infections in early life in areas where multiple flaviviruses co-circulate, particularly considering the immune interactions between ZIKV and dengue and the future possibility of ZIKV vaccination in women of childbearing potential. This study also shows the benefits of cord blood sampling for serologic surveillance of infectious diseases in resource-limited settings.

Keywords: Zika; antibodies; antibody-dependent enhancement; flavivirus; humoral immunity; neonatal immunity; passive immunity.

Figure 1

Seroprevalence of ZIKV-reactive IgG in cord blood. ZIKV IgG (A) and ZIKV EDIII (B) ELISA were performed on cord blood plasma diluted 1:200 from Cohort 1 (n = 104) and Cohort 2 (n = 135). The cutoff for positivity was determined on each plate by running negative control plasma. The percent seropositivity for each assay is shown beneath the cohort names, and the geometric mean (GM) of OD for positive samples is shown beneath that. (C) The OD value from the ZIKV EDIII ELISA for available paired samples (n = 102) from mother peripheral blood and cord blood are graphed in a scatter plot to assess the efficiency of ZIKV-reactive IgG transfer across the placenta. Pearson correlation analysis was performed, and the R² and p values are included in the upper left corner of each graph. ns, not significant; ****, p < 0.0001 according to an unpaired Student’s t-test comparing differences between GM of Cohort 1 vs. Cohort 2.

Figure 2

ZIKV-reactive antibody kinetics and IgG subtype. Decay of IgG reactive to ZIKV (A) and ZIKV EDIII (B) is shown for infants in both PW1 and PW2 cohorts. Spaghetti plots of trajectories were generated from the OD from each ELISA as a ratio to the LLOD, using a continuous time scale of weeks of life for each infant. Data from individual infants are shown as semi-transparent lines. Fit lines summarizing all data were calculated using an exponential decay (red dashed line) and a quadratic polynomial (blue dashed line) model. (C) IgG3 reactivity to ZIKV (left panel) and ZIKV EDIII (right panel) is shown in a subset of cord blood samples (n = 10 per subgroup) from subgroups indicated on the x-axis. OD, optical density; LLOD, lower limit of detection. X-axis legend of left graph of panel C: needs space between LOW and EDIII in the 2 group from left.

Figure 3

Incident flavivirus infection detected after initial decay of maternal-derived anti-ZIKV IgG in early life. Paired samples available from time points before and after the observed increase in anti-ZIKV IgG were selected for NAb testing. (A) Representative raw FRNT data measuring NAb to ZIKV (left) and DENV2 (right) are shown for three subjects. (B,C) eFRNT50 values for ZIKV (left) and DENV2 (right) are shown for suspected cases of incident flavivirus infection in Cohort 1 (B) and Cohort 2 (C). For nine samples with suspected incident flavivirus infection., and eFRNT50 values for paired samples are linked by a solid line.